基于HSI空间模糊聚类的彩色图像分割新方法  

Fuzzy clustering based on HIS color space for color image segmentation

作　　者：邱磊[1] 管一弘[1] 

机构地区：[1]昆明理工大学理学院,云南昆明650093

出　　处：《微型机与应用》2011年第22期46-48,共3页Microcomputer & Its Applications

摘　　要：模糊C均值(FCM)被广泛应用于彩色图像分割中,但传统的模糊C均值由于没有考虑空间信息,因此对噪声特别敏感。针对此问题,提出了一种在HIS颜色空间结合像素邻域空间信息的模糊聚类新方法。实验结果表明,此方法对高噪声图像有较好的处理结果。Fuzzy C-Mean（FCM） algorithm has been wildely used in the field of color image segmentation. But the convention- al FCM is sensitive to the noise because the spatial information is not being considered. To solve this problem, a new fuzzy clustering algorithm in HIS color space is proposed based on pixel＇s spatial neighborhood and pixel value. Experimental results prove that it has a good result to high noise image.

关 键 词：彩色图像分割 HIS彩色空间 邻域均值 模糊C均值 

分 类 号：TN911.7[电子电信—通信与信息系统]